Accuracy-based optimization of the filtered Finite-Difference Time-Domain method

The amount of the high frequency components removed by the low-pass filter has significant influence on the accuracy and the upper limit of NCFL for the stable computation. So far there is little report on the accuracy-based parameter setting for the filtered FDTD method. In the conference, we present our study on the optimisation of the cut-off frequency of the low-pass filter from the viewpoint of maximum NCFL and the accuracy as well as the computational efficiency.